A level-based learning swarm optimizer for large-scale optimization

Qiang Yang, Wei-Neng Chen, Jeremiah Da Deng, Yun Li, Tianlong Gu, Jun Zhang

Research output: Contribution to journalArticle

  • 1 Citations

Abstract

In pedagogy, teachers usually separate mixed-level students into different levels, treat them differently and teach them in accordance with their cognitive and learning abilities. Inspired from this idea, we consider particles in the swarm as mixed-level students and propose a level-based learning swarm optimizer (LLSO) to settle large-scale optimization, which is still considerably challenging in evolutionary computation. At first, a level-based learning strategy is introduced, which separates particles into a number of levels according to their fitness values and treats particles in different levels differently. Then, a new exemplar selection strategy is designed to randomly select two predominant particles from two different higher levels in the current swarm to guide the learning of particles. The cooperation between these two strategies could afford great diversity enhancement for the optimizer. Further, the exploration and exploitation abilities of the optimizer are analyzed both theoretically and empirically in comparison with two popular particle swarm optimizers. Extensive comparisons with several state-of-the-art algorithms on two widely used sets of large-scale benchmark functions confirm the competitive performance of the proposed optimizer in both solution quality and computational efficiency. Finally, comparison experiments on problems with dimensionality increasing from 200 to 2000 further substantiate the good scalability of the developed optimizer.

LanguageEnglish
Pages578-594
Number of pages17
JournalIEEE Transactions on Evolutionary Computation
Volume22
Issue number4
Early online date5 Sep 2017
DOIs
StatePublished - 31 Aug 2018

Fingerprint

Large-scale Optimization
Swarm
Students
Computational efficiency
Evolutionary algorithms
Scalability
Particle Swarm Optimizer
Pedagogy
Learning Strategies
Evolutionary Computation
Experiments
Computational Efficiency
Exploitation
Fitness
Dimensionality
Enhancement
Learning
Benchmark
Experiment

Keywords

  • exemplar selection
  • high-dimensional problems
  • large-scale optimization
  • level-based learning swarm optimizer (LLSO)
  • particle swarm optimization (PSO)

Cite this

Yang, Qiang ; Chen, Wei-Neng ; Deng, Jeremiah Da ; Li, Yun ; Gu, Tianlong ; Zhang, Jun. / A level-based learning swarm optimizer for large-scale optimization. In: IEEE Transactions on Evolutionary Computation. 2018 ; Vol. 22, No. 4. pp. 578-594
@article{4ebfda6f2c1b45ce8f5257c2e1d96c41,
title = "A level-based learning swarm optimizer for large-scale optimization",
abstract = "In pedagogy, teachers usually separate mixed-level students into different levels, treat them differently and teach them in accordance with their cognitive and learning abilities. Inspired from this idea, we consider particles in the swarm as mixed-level students and propose a level-based learning swarm optimizer (LLSO) to settle large-scale optimization, which is still considerably challenging in evolutionary computation. At first, a level-based learning strategy is introduced, which separates particles into a number of levels according to their fitness values and treats particles in different levels differently. Then, a new exemplar selection strategy is designed to randomly select two predominant particles from two different higher levels in the current swarm to guide the learning of particles. The cooperation between these two strategies could afford great diversity enhancement for the optimizer. Further, the exploration and exploitation abilities of the optimizer are analyzed both theoretically and empirically in comparison with two popular particle swarm optimizers. Extensive comparisons with several state-of-the-art algorithms on two widely used sets of large-scale benchmark functions confirm the competitive performance of the proposed optimizer in both solution quality and computational efficiency. Finally, comparison experiments on problems with dimensionality increasing from 200 to 2000 further substantiate the good scalability of the developed optimizer.",
keywords = "exemplar selection, high-dimensional problems, large-scale optimization, level-based learning swarm optimizer (LLSO), particle swarm optimization (PSO)",
author = "Qiang Yang and Wei-Neng Chen and Deng, {Jeremiah Da} and Yun Li and Tianlong Gu and Jun Zhang",
year = "2018",
month = "8",
day = "31",
doi = "10.1109/TEVC.2017.2743016",
language = "English",
volume = "22",
pages = "578--594",
journal = "IEEE Transactions on Evolutionary Computation",
issn = "1089-778X",
number = "4",

}

A level-based learning swarm optimizer for large-scale optimization. / Yang, Qiang; Chen, Wei-Neng; Deng, Jeremiah Da; Li, Yun; Gu, Tianlong; Zhang, Jun.

In: IEEE Transactions on Evolutionary Computation, Vol. 22, No. 4, 31.08.2018, p. 578-594.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A level-based learning swarm optimizer for large-scale optimization

AU - Yang,Qiang

AU - Chen,Wei-Neng

AU - Deng,Jeremiah Da

AU - Li,Yun

AU - Gu,Tianlong

AU - Zhang,Jun

PY - 2018/8/31

Y1 - 2018/8/31

N2 - In pedagogy, teachers usually separate mixed-level students into different levels, treat them differently and teach them in accordance with their cognitive and learning abilities. Inspired from this idea, we consider particles in the swarm as mixed-level students and propose a level-based learning swarm optimizer (LLSO) to settle large-scale optimization, which is still considerably challenging in evolutionary computation. At first, a level-based learning strategy is introduced, which separates particles into a number of levels according to their fitness values and treats particles in different levels differently. Then, a new exemplar selection strategy is designed to randomly select two predominant particles from two different higher levels in the current swarm to guide the learning of particles. The cooperation between these two strategies could afford great diversity enhancement for the optimizer. Further, the exploration and exploitation abilities of the optimizer are analyzed both theoretically and empirically in comparison with two popular particle swarm optimizers. Extensive comparisons with several state-of-the-art algorithms on two widely used sets of large-scale benchmark functions confirm the competitive performance of the proposed optimizer in both solution quality and computational efficiency. Finally, comparison experiments on problems with dimensionality increasing from 200 to 2000 further substantiate the good scalability of the developed optimizer.

AB - In pedagogy, teachers usually separate mixed-level students into different levels, treat them differently and teach them in accordance with their cognitive and learning abilities. Inspired from this idea, we consider particles in the swarm as mixed-level students and propose a level-based learning swarm optimizer (LLSO) to settle large-scale optimization, which is still considerably challenging in evolutionary computation. At first, a level-based learning strategy is introduced, which separates particles into a number of levels according to their fitness values and treats particles in different levels differently. Then, a new exemplar selection strategy is designed to randomly select two predominant particles from two different higher levels in the current swarm to guide the learning of particles. The cooperation between these two strategies could afford great diversity enhancement for the optimizer. Further, the exploration and exploitation abilities of the optimizer are analyzed both theoretically and empirically in comparison with two popular particle swarm optimizers. Extensive comparisons with several state-of-the-art algorithms on two widely used sets of large-scale benchmark functions confirm the competitive performance of the proposed optimizer in both solution quality and computational efficiency. Finally, comparison experiments on problems with dimensionality increasing from 200 to 2000 further substantiate the good scalability of the developed optimizer.

KW - exemplar selection

KW - high-dimensional problems

KW - large-scale optimization

KW - level-based learning swarm optimizer (LLSO)

KW - particle swarm optimization (PSO)

UR - http://www.scopus.com/inward/record.url?scp=85029151806&partnerID=8YFLogxK

UR - https://ieeexplore.ieee.org/xpl/tocresult.jsp?isnumber=8423223

U2 - 10.1109/TEVC.2017.2743016

DO - 10.1109/TEVC.2017.2743016

M3 - Article

VL - 22

SP - 578

EP - 594

JO - IEEE Transactions on Evolutionary Computation

T2 - IEEE Transactions on Evolutionary Computation

JF - IEEE Transactions on Evolutionary Computation

SN - 1089-778X

IS - 4

ER -